Predicting Cloud Computing Adoption in IoMT Using Deep Learning Approach

Abstract

Cloud computing permits appropriate, on demand computer network structure connections to a shared group of prearranged computing resources, including computer hybrid system, supercomputer, data storage system, application software procedures, and computer amenities. Through the IT industry and other business organizations the primary healthcare management and services can be greatly improved and benefitted from the cloud computing assessment as a superior procedure of IT outsourcing. It has been demonstrated that there is a connection among cloud computing and Internet of Things (IoT) outsourcing. The IoT paradigm shift has contributed to current distributed computing systems research and provided a data source for a variety of scientists and students use. The Internet of Medical Things (IoMT), a variant of IoT, solely supports health IT outsourcing based on data invasion and the development of innovative technologies, especially, in the healthcare sector, as this would not exempt the medical profession from this trend. The use of a qualitative approach to analyse the interaction between cloud computing and IoMT is what makes this article significant. The IoMT, which supports health IT outsourcing cantered on data invasion and the advancement in the new technologies as regards the healthcare sector, does not exempt the medical profession from this trend. The use of a qualitative approach to analyse the interaction between cloud computing and IoMT is what makes this article significant. The adoption of cloud computing in IoMT was also predicted in the research adopting a deep learning approach. From the analysis, it confirms that there is a notable relationship existing concerning the cloud computing and the IoMT. In the same manner, the adoption of cloud computing at IoMT was anticipated by the established deep learning methodology using an enhanced sigmoid transfer neural network. The findings demonstrated that the deep learning model out performed other similar models with true positive, false negative, precision, recall, F1 score, and accuracy values of 95.6%, 4.5%, 95.7%, 95.6%, and 95.545%, respectively.